Wrist Band Photoplethysmography Autocorrelation Analysis Enables Detection of Atrial Fibrillation Without Pulse Detection

Abstract

Atrial fibrillation is often asymptomatic and intermittent making its detection challenging. A photoplethysmography (PPG) provides a promising option for atrial fibrillation detection. However, the shapes of pulse waves vary in atrial fibrillation decreasing pulse and atrial fibrillation detection accuracy. This study evaluated ten robust photoplethysmography features for detection of atrial fibrillation. The study was a national multi-center clinical study in Finland and the data were combined from two broader research projects (NCT03721601, URL: https://clinicaltrials.gov/ct2/show/NCT03721601 and NCT03753139, URL: https://clinicaltrials.gov/ct2/show/NCT03753139). A photoplethysmography signal was recorded with a wrist band. Five pulse interval variability, four amplitude features and a novel autocorrelation-based morphology feature were calculated and evaluated independently as predictors of atrial fibrillation. A multivariate predictor model including only the most significant features was established. The models were 10-fold cross-validated. 359 patients were included in the study (atrial fibrillation n = 169, sinus rhythm n = 190). The autocorrelation univariate predictor model detected atrial fibrillation with the highest area under receiver operating characteristic curve (AUC) value of 0.982 (sensitivity 95.1%, specificity 93.7%). Autocorrelation was also the most significant individual feature (p < 0.00001) in the multivariate predictor model, detecting atrial fibrillation with AUC of 0.993 (sensitivity 96.4%, specificity 96.3%). Our results demonstrated that the autocorrelation independently detects atrial fibrillation reliably without the need of pulse detection. Combining pulse wave morphology-based features such as autocorrelation with information from pulse-interval variability it is possible to detect atrial fibrillation with high accuracy with a commercial wrist band. Photoplethysmography wrist bands accompanied with atrial fibrillation detection algorithms utilizing autocorrelation could provide a computationally very effective and reliable wearable monitoring method in screening of atrial fibrillation.

Keywords: algorithms; arrhythmia detection; atrial fibrillation; atrial fibrillation detection; autocorrelation; photoplethysmography; pulse detection; stroke.

FIGURE 1

Standards for Reporting Diagnostic Accuracy Studies (STARD) flow diagram of the study patient flow. A total of 555 patients were screened in the participating hospitals KUH, HUS, and NKCH. 359 patients were included in the analysis. AF, atrial fibrillation; ECG, electrocardiogram; SR, sinus rhythm; PPG, photoplethysmography; RBBB, right bundle branch block.

FIGURE 2

Example recordings. PPG (upper) and ECG (lower) recordings from three patients. Panel (A) shows a patient with sinus rhythm, panel (B) atrial fibrillation with lenient heart rate and panel (C) atrial fibrillation with high heart rate. Algorithm ECG QRS detection points and PPG pulse detection points are marked with red circles. A PIN time series was formed with detected PPG pulses for PIN-based AF detection features. ECG, electrocardiogram; PPG, photoplethysmography; HR, heart rate.

FIGURE 3

Autocorrelation. PPG (upper) and ECG (lower) recordings from a patient with sinus rhythm (A1) and atrial fibrillation (B1). Corresponding autocorrelation values were calculated for 1-min samples of PPG signal for each patient. First 10 s of example recordings and calculated autocorrelation values (A2 and B2) are shown in panels. Autocorrelation is a feature calculated straight from the signal and it requires no pulse detection. It is the correlation between a signal and its delayed copy as a function of delay. ECG, electrocardiogram; PPG, photoplethysmography.

FIGURE 4

Averaged AF detection ROC curve of the univariate models and the multivariate predictor model.